Diabetes mellitus is a growing problem worldwide. In the US it affects over 18 million people and results in annual health care costs exceeding $130 billion. Insulin therapy of Type 1 diabetes, and in advanced cases of Type 2 diabetes, does not prevent serious long-term complications including neuropathy, vascular disease, retinopathy and renal failure. Transplantation of pancreatic islets to restore insulin production offers significant promise. However, the supply of donor pancreata falls far short of meeting the medical need. New sources of insulin producing cells will be required to realize the full potential of cell therapy for diabetes. We propose to generate pancreatic beta lineage cells by in vitro differentiation of stem cells isolated from amniotic fluid. These "AFS cells" are capable of both extensive expansion and differentiation into derivatives of all three embryonic germ layers. Our Preliminary Studies showed that mouse AFS cells can yield insulin producing cells and islet-like cell clusters ("neo-islets"), promoted by expression of the pancreatic transcription factor PDX-1. We now propose to produce neo-islets from human and non-human primate (NHP) AFS cells. To efficiently generate insulin producing cells, we will optimize delivery of a plasmid vector to express PDX-1, and will systematically test growth factors and substrates shown previously to promote pancreatic beta cell differentiation. The resulting neo-islets will be compared with authentic human and NHP pancreatic islets using tests developed for clinical transplantation. To assess their ability to restore control of glucose metabolism and production of insulin and C-peptide, neo-islets will be implanted in immunodeficient mice made diabetic with streptozotocin (STZ). Transplantation in STZ-treated NHP will assess the function of neo-islets in a model physiologically more similar to humans. NHP AFS cell lines will be derived after amniocentesis of pregnant mothers. These stem cells will be used to generate neo-islets for autologous transplantation into the corresponding offspring. The same donor cells will be compared in allogeneic recipients using clinically relevant immunosuppression regimens. Successful development of an abundant source of transplantable insulin producing cells potentially would have a profound impact on the treatment of a major public health problem.